Mexican Exposures
Elizabeth F. S Roberts
Senior Research Proposal:
Submitted to the National Science Foundation
For co-review - Cultural Anthropology/STS
Project P.I. Elizabeth F.S. Roberts
Mexican Exposures:
A Bio-ethnography of Six Urban Families
I. OVERVIEW
In recent decades, social and biological scientists have found twentieth-century distinctions between nature and culture increasingly inadequate for describing the reality of human life. Yet while researchers are pursuing ways to transcend the nature/culture dichotomy, there have been few cross-domain attempts to apply these insights to research on enduring problems like the relationship between life circumstances and disease. Medical anthropology has long been focused on this relationship, but there are at least two challenges for medical anthropological ethnographic research that moves beyond the nature/culture divide . One is the epistemological and pragmatic difficulty of symmetrically combining biological with ethnographic data without making a priori assignments of phenomena to either nature or culture, which tends to give nature more weight. A second issue is the question of how to deploy insights from science and technology studies (STS) that both cultural and biological processes are constructed and contingent. Meeting these challenges creates the prospect of a medical anthropology in which biological research is both a source of data and an object of study. The proposed project takes up both provocations with a two phase, multi-sited, and interdisciplinary study of the interactions of bodies, environments, and life circumstances that aims to forge new models for reflexive bio-ethnographic research.
Mexican Exposures involves a three-year collaboration between myself (a medical anthropologist), and a team of environmental health researchers at the University of Michigan. It seeks to develop a bio-ethnographic approach that integrates biological and ethnographic data about the larger histories and life circumstances that shape disease. Since 1995, the birth-cohort study in Mexico City called ELEMENT (Early Life Exposure in Mexico to Environmental
Toxicants) has studied the long-term and intergenerational physiological and neurological effects of in utero and early childhood exposure to chemicals like lead and mercury. ELEMENT researchers have begun to link these exposures to obesity, diabetes, and attention-deficit hyperactivity disorder (ADHD). While ELEMENT has made several key findings about chemical exposures, its approach has tended to situate key disease-transmission mechanisms inside individual bodies rather than within larger historical and economic processes. My effort involves building on and expanding correlations between individual health and environment through the inclusion of data about the specific historical, social and material contexts that shape ELEMENT participants’ lives, as well as data on how participants’ lives are shaped by ELEMENT research.
The project has two phases: first, a year-long ethnography of six ELEMENT participant families living in poor and working-class neighborhoods, focusing on household and neighborhood environments and histories relevant to the production of bodily states; second, a two-year phase in which ELEMENT researchers and I combine our biological and ethnographic exposure data to produce a multifaceted and more complex account of the links between ill-health and life circumstances. In both these phases, I will also investigate the scientific process itself: how participants’ lives are shaped by research (including my own), and how ELEMENT researchers attempt to universalize the knowledge gleaned from monitoring individual participants’ bodies.
1

Mexican Exposures
Elizabeth F. S Roberts
II. RESEARCH QUESTIONS AND SPECIFC AIMS
By adopting a symmetrical analysis of the biological and historical construction of locally situated bodies, Mexican Exposures will address the following sets of questions:
1) What life circumstances and material conditions affect bodily states? That is, what effects do factors such as economic conditions, urban infrastructure, religious practice, and kinship relations play in exposure to toxins? What other phenomena might interact with toxins or influence bodily states such as obesity, diabetes, and ADHD? How do research participants themselves understand the forces that shape their well-being? These questions will be addressed through ethnographic research among ELEMENT participant families.
2) How is scientific knowledge constructed? STS approaches view expert knowledge itself as an object of study, a phenomenon that affects the dynamics being studied. How is generalizable “universal” scientific knowledge produced from concrete, local data from specific bodies? How do local data and scientific findings circulate? What are the effects of this process on participants? How do they interpret their experience of participation in the research? My preliminary research with the ELEMENT project suggests that participation in ELEMENT may produce looping effects on the participants’ life circumstances and conditions.
3 ) How can ethnographic and biological data be combined in accounts of bodily states?
My preliminary research suggests some directions for this exploration. This project requires refashioning received categories, such as the concept of exposure and the nature/culture divide, and will draw on my ethnographic research, ELEMENT data, existing theoretical literature, and collaboration with ELEMENT scientists.
These questions are correlated with Mexican Exposures’ three specific aims.
1) To collect data on the environments and circumstances that shape complex states such as obesity and diabetes through ethnographic household research.
2) To study the production and effects of scientific—and ethnographic—research associated with ELEMENT’s project and integrate the findings into a larger analysis of environments, exposure, and bodily states in Mexico City.
3) To develop a bio-ethnographic approach that will expand the concept of exposure by incorporating both biological data (from the ELEMENT team) and my ethnographic data, producing novel cross-disciplinary accounts of the connections between environment, life circumstances, and health.
III. BACKGROUND LITERATURE
The starting point for my project is the critique of an epistemological and ontological divide between nature and culture. This critique provides the potential for new explorations of the interaction between environments, bodies, and health. Its emergence in social science and area studies parallels directions in recent biological scholarship. Mexican Exposures builds on four interrelated anthropological and social-scientific literatures: (1) the examination of nature/culture distinctions in the biological sciences, anthropology, and STS; (2) longstanding anthropological and public policy debates about the “culture of poverty”; (3) Latin American anthropology and history focusing on the malleability of nature; and (4) the medical anthropological and STS literature on the political economy of biomedicine and science, including the ways scientific knowledge is constructed through the recruitment of particular subject populations.
(1) The practice of assigning phenomena to the discrete categories of nature or culture is quite recent, emerging only in the late nineteenth century in Western Europe and the United States
(Daston and Galison 2007; Degler 1991; Fox Keller 2010). Previously, humans were understood as being shaped by the environment in ways we now think of as Lamarckian (i.e., through the inheritance of acquired characteristics). Turn-of-the-century biological scientists, like August
Weismann, developed models of the germ line that deemphasized the organism’s interaction with the environment, endowing new entities (first germ lines and then genes) with individual and
“particulate” powers that allowed them to exist and perpetuate themselves unaffected by external
2

Mexican Exposures
Elizabeth F. S Roberts conditions (Fox Keller 2010). This divide produced an academic division of labor whereby the natural sciences came to examine the “hard” objects of nature, and the social sciences investigated malleable phenomena like culture.
These divisions between nature and culture informed debates about whether characteristics of race, sex, health, and ability are intrinsic or malleable, and raising questions about whether problems like disease and persistent poverty could ever be alleviated by socialwelfare policies and programs (Mead 1953; Rabinow 1996a; Stepan 1991). In their insistence on fixity, North American eugenics arguments of the early twentieth century were similar to later socio-biological arguments that argued for DNA as an unchanging blueprint for physiology and behavior that operated largely independently of the social environment. These “hard” versions of nature allowed for a vast array of claims about types (“criminals,” “immigrants,” “imbeciles”) and genes that determined everything from race, sexuality, and intelligence to the incidence of poverty and disease. Arguing back, social scientists, especially midcentury cultural anthropologists, positioned themselves against biological determinism, asserting the power of culture in shaping human life and difference (Boas 1922; Degler 1991; Mead 1953). At the same time, cultural anthropologists argued for a “universal” biology that challenged nineteenth- and twentieth-century ideologies of innate racial difference (Baker and Patterson 1994). However, the assumption of a universal biology underlying multiple local cultures made biology appear inviolable by social circumstances, and culture seem primarily ideational, without any ability to shape bodies beyond surface appearances (Boas 1922; Kroeber 1917).
In recent years this model has been challenged within both the biological and social sciences. As genome sequencing technology and projects like the Human Genome Project revealed the parsimony of the human genome, these claims began to lose force, and scientists began to look elsewhere to explain the complexity of bodies and behaviors (Cloud 2010; Reardon
2005). Epigenetics, a term coined in the 1940s, has emerged recently in environmental health, microbiology, toxicology, oncology, and psychology in response to the inability of genetics to provide full explanations for phenotype and behavior. It offers a framework for the study of the ways external environmental exposures or “triggers,” such as pollutants and foods, can alter the expression of genes and exert long-term effects on phenotype, even in adult bodies, brains, and behavior, through mechanisms such as DNA methylation, histone modifications, and DNAbinding proteins (Hewagama and Richardson 2009; Weaver, et al. 2004).
Epigenetic researchers define exposure broadly, to include such factors as climate, stress, and nutrition. The complex set of interactions between external environmental exposure and the maternal body, and the complexity of epigenetic mechanisms and effects (which can vary by individual, by organ, and by cell), have challenged the fundamental idea of a divide between the domains of nature and culture, or organism and environment (Szyf, et al. 2005; Waterland and
Michels 2007). Given their discipline’s focus on examining relationships between environments and organisms, environmental health scientists, such as ELEMENT researchers, are especially drawn to the tools of epigenetics.
In the social sciences in the last quarter century, feminist theorists, philosophers, historians, sociologists, and anthropologists of science have increasingly found the nature/culture divide inadequate for describing the reality of human life. Some social science scholars have moved away from social constructionism, which posits that social ideas construct our sense of the material world, towards constructionism, which sees everything as constructed, even nature and the material world (Haraway 1991; Latour 2005; Mol 2002). Whereas social constructionism distinguishes between the ideational social world and the object world, constructionists insist instead on the lack of separation between the social and the material world and see both as malleable. In other words, both objects and facts are real and constructed (Latour 2010). These scholars began to argue for the material interconnectedness of the world, proposing terms like
“nature/cultures” (Haraway 1991) and “biosocialities” (Rabinow 1996b), terms that analyze the world without privileging natural phenomena.
3

Mexican Exposures
Elizabeth F. S Roberts
A key constructionist concept that has emerged in medical anthropology has been that of
“local biologies”(Lock 1993). As developed by Margaret Lock, the concept emphasizes how
“biological and social processes are inseparably entangled over time, resulting in human biological difference” (Lock and Nguyen 2010, 90). In this view, biological information
“produces a real but partial picture” of human life (Lock and Nguyen 2010, 93). A “local biologies” approach includes biological processes and biological data, while also understanding them as historically contingent, part of the “looping effects” that condition life circumstances. So far, however, few local-biological accounts have made biological data intrinsic to their analysis
(for exceptions see Fullwiley 2011; Lock 1993). In developing an analytic method to examine biological and social exposures, I seek to realize these theoretical appeals for a constructed, localbiological account of the world.
The current focus in the social sciences on the ontological malleability of biology resonates with trends in the biosciences that posit complex conditioning entanglements between organisms and environment—especially environmental health science, evolutionary developmental biology, and epigenetics (Barker 1990; Bateson, et al. 2004; Jirtle and Skinner 2007; Suter, et al. 2010).
Anthropology and STS have much to offer biological researchers, however, who tend to narrowly define what they call “social determinants” and continue to situate key disease-transmission mechanisms exclusively inside individual bodies rather than within larger histories (Lock 2013).
For example, epigenetic health researchers in the United States have tended to direct the implications of their findings towards individuals instead of larger infrastructural problems that shape disease. In particular, this literature risks regarding women as reproductive “containers” wholly responsible for the health of the fetus, whose bodies therefore come to be seen as the appropriate sites for individual intervention and control (Szyf, et al. 2005; Weaver, et al. 2004).
Mexican Exposures will be well positioned to provide a critique of individualizing and socially deterministic accounts of disease.
(2) The “culture of poverty” concept arose in anthropology as a means to avoid deterministic biological and racial explanations for ongoing poverty, but ultimately it produced another form of determinism. The concept was employed in contentious U.S. anthropology, social science and public-policy debates in the 1960s to attribute lack of economic advancement to the
“cultural mentalities” of the poor. A key early text used, or misused, in the popularization of this concept was Oscar Lewis’s Five Families: Mexican Case Studies in the Culture of Poverty
(Lewis 1975), a much-criticized monograph that described one day in the life of five Mexican families in the 1950s, four of them in Mexico City. If we read it through a contemporary lens, however, it’s difficult to see how such a reductive and deterministic notion of “culture” was ever taken from Five Families.
The book’s analysis is based on material contingency, with its focus on how water delivery, corn grinding and tortilla making, urban architecture, migration, government policy, and the lack of employment possibilities organize the daily practices and affective lives of family members. Nonetheless, the earlier reductive reading of Lewis’s work as about “culture” was influential. The uptake of this classic ethnographic account of poverty is instructive today, as some researchers in public policy, sociology, and even epigenetics are again attracted to the
“culture of poverty” concept, remaining mostly uncritical of how the concept separates culture from the material conditions and forms of dominance that produce inequality (Bourgois 2001;
Small et al. 2010).
Given their insistence that culture cannot be separated from material and biological conditions, scholars in medical anthropology and STS are well positioned to problematize the notion that culture is an impediment to economic advancement or can be analyzed as a “variable.”
In recent decades, anthropologists and sociologists have produced nuanced accounts of the conditions that perpetuate inequality (Bourgois 1995; Edin and Lein 1997; Katz 2013; Mullings and Wali 2001). These scholars explore the interrelations of poverty and environment without rooting poverty in either the bodies or the mentalities of the poor. Nonetheless, in their parsing of the conditions that perpetuate poverty and ill-health, these studies leave physiological processes
4

Mexican Exposures
Elizabeth F. S Roberts unexamined. This lacuna points to the need for reflexive awareness of the possibilities for combined biological and ethnographic findings—or bio-ethnographic research—to provide symmetrical accounts of life circumstances like poverty without deploying biology or culture as their locus. In effect, Mexican Exposures anticipates providing a successor for Five Families written from a bio-ethnographic perspective.
(3) Current biological models of the entanglement of environment and organism have some similarities with the longstanding Latin American emphasis on the reciprocal malleability of bodies and environments. In Latin America, more than in North America, the body is seen and experienced not as fixed but as plastic and cultivable. This view has often been characterized by
North American scholars as Lamarckian (Cadena 2000; Stepan 1991). Mexican Exposures is thoroughly engaged with the anthropology and history of Latin America, particularly work that examines mestizaje (race mixture), state formation, and economic policy in relation to nature, and links racial and other bodily states to education, occupation, and health (Blum 2009; Cadena
1995; Caldeira 2000; Colloredo-Mansfeld 1998; Mallon 1996; Pitt-Rivers 1973; Wade 1993;
Zulawski 2007). These themes have been taken up in recent years by scholars concerned with
Latin American bioscientific research efforts concerning nutrition, ecology, genetics, and pharmaceuticals (Brotherton 2012; Cueto 2001; Gibbon 2013; Hartigan 2013; Hayden 2003;
Hayden 2007; Laveaga 2009; Nading 2013; Whitmarsh 2008; Yates-Doerr 2012). Bringing the
Latin American area literature about the malleability of nature into critical conversation with the
STS literature dislodges current assumptions about the originality of current constructionist STS models of bodily plasticity and enables a close examination of the ways ELEMENT participants themselves model entanglements of ill-health and life conditions, which I discuss below.
(4) In addition to building on the Latin America area studies literature that complicates nature/culture distinctions, Mexican Exposures will engage with the STS and anthropological literature that parses the “looping” relationship between the way political and economic conditions produce particular biomedical subject populations and the way these populations in turn shape scientific knowledge production (Hacking 1995; Lock and Nguyen 2010). Scholars who examine the globalized pharmaceutical and bioscientific apparatus have found that the prevalent biomedical assumption of apolitical, universal bodies and biological processes actively shapes the results of clinical trials and policy interventions (Biehl 2005; Biehl and Petryna 2001;
Briggs and Mantini-Briggs 2003; Montoya 2007; Montoya 2013; Nguyen 2005; Petryna 2009;
Rose and Novas 2005). These assumptions obscure the material differences that make certain groups into test subject populations and others into intended patient populations. Ignoring how different histories of poverty, nutrition, access to resources, and previous pharmaceutical exposure intensively shape data collection and analysis has had detrimental effects on knowledge production and on the bodies of both subject populations and patients (Das 2006; Lakoff 2005;
Maternowska 2006; Petryna, et al. 2006). As I discuss below, my pilot observations in Mexico
City last summer led me to believe that similar processes may be at work in the ELEMENT project, which makes it imperative to analyze how both ELEMENT and my ethnographic research shape ELEMENT participants’ lives.
The analytical and empirical directions in medical and cultural anthropology, public policy, STS, Latin American studies, and cutting-edge epigenetics research raise challenges for the ethnographic study of bodies and health. What would a symmetrical analysis of biology and social life look like in an examination of particular embodied experiences? How could we view both research subjects and scientific research itself as interacting elements in the same field?
What methods and frameworks would allow a medical anthropologist to combine biological data with ethnographic materials without privileging the biological?
IV. RESEARCH CONCEPTS
Building on these analytical currents, the proposed project develops a model for empirical and analytical approaches that combine biological and ethnographic data, including
5

Mexican Exposures
Elizabeth F. S Roberts reflection on knowledge construction, in what I call a bio-ethnographic approach . The research brings such an approach into engagement with a concept central to environmental health studies, exposure .
In collaboration with ELEMENT researchers, my project will investigate the production of diabetes, obesity, and possibly ADHD, in Mexico City by associating separately derived ethnographic and biological data in order to analyze the relationship of life circumstances and disease. Taking cues from STS, my approach will insist on the symmetrical analysis of both kinds of data, without privileging either one as more real or fixed.
A bio-ethnographic approach might sound similar to what American biological anthropologists have formulated in recent decades as a “bio-cultural synthesis,” which explores the role of culture in shaping human biology and behavior, but there are important differences.
Bio-cultural anthropologists have contributed to the understanding of a wide range of phenomena, like HIV transmission, high-altitude adaptation, and lactase persistence (Beall 2006; Brabec, et al.
2007; Hadley, et al. 2010; Lindstrom, et al. 2011; Wiley 2008). By maintaining culture as distinct from biology, however, the approach yields a synthesis that remains asymmetrical. It leaves unexamined the historical and economic conditions that continuously shape biological processes and scientific study itself. In effect, this reification of culture as separate from biology is similar to social constructionism, which posits an object world separate from the ideational social world.
By proposing a bio-ethnographic account, I am not simply adding culture to biology in order to arrive at a better understanding of human biological adaptation, but instead combining two different kinds of methodologies—ethnographic observation and biological sampling—in a synthetic symmetrical analysis that understands environment-body interactions as always relational and constructed phenomena.
At this early stage, I am imagining exposure as a crucial conceptual term for bringing these two methods together. In the biological sciences, the noun exposure tends to have a specific and negative meaning, referring to contact with harmful phenomena, such as exposure to lead. In other contexts, however, exposure can refer more broadly to neutral or positive encounters with external phenomena, such as exposure to the arts or foreign foods.
Some social science research on environmental health science has adopted the concept of exposure as a means to bring together the study of health and of the environment, two realms that have long been examined separately (Brown, et al. 2011; Mitman, et al. 2004; Shostak 2013).
However, in this context, exposure is somewhat taken for granted as an agreed-upon technical term transparently describing events that happen to bodies in environments. I propose documenting ethnographically how exposure is defined by ELEMENT scientists, study participants, and myself, and then examining exposure as a set of processes that include historical and political-economic dimensions, which are made visible through scientific procedures and infrastructures.
Biological researchers are also beginning to move away from considering exposures as discrete events, instead considering organisms as being in a dynamic, interactive relationship with the environment. Within this relational model, exposures to toxins are understood less as isolated experiences with singular effects than as part of a “complex web of ever-changing interactions”
(Jablonka and Lamb 2005, 239). Exposures are now seen as ongoing “conditioning media” that persist within organisms, reshaping them in ways that then shape those media’s future uptake within organisms (Landecker 2011). Mexican Exposures deploys exposure as a rubric to facilitate a bio-ethnographic approach that combines environmental health and critical STS and anthropological frameworks for understanding the construction of bodily states.
V. RESEARCH SETTING AND PRELIMINARY RESEARCH
In 1992, the United Nations declared Mexico City the most polluted city on earth. In light of this designation, a team of U.S.-based environmental health researchers began a collaborative project with public health officials in Mexico, named ELEMENT, to research the effects of
6

Mexican Exposures
Elizabeth F. S Roberts pollution and chemical exposures, particularly lead, on fetal and childhood growth and neurological development. The intertwined geographic, environmental, political, economic, and historical conditions of Mexico City have provided ELEMENT with an ideal site to study the effects of chemical exposure on development. Mexico City lies in a dry lakebed 7,350 feet above sea level. Intense solar radiation at this altitude worsens air pollution. Since 1992, public policy has mitigated air pollution by banning lead in gasoline and certain kinds of catalytic converters in cars. At the same time, obesity rates have risen dramatically, resulting in Mexico’s being designated the fattest industrial nation in 2013. The rise in obesity and diabetes has put increased pressure on Mexico’s health care system, Seguro Social. The system is notorious for its long waits, brusque and impersonal care, and lack of drugs and services. This complex set of circumstances creates relevant project conditions and a willing and compliant subject population for ELEMENT, who through their participation receive attention unavailable through Seguro
Social.
ELEMENT, funded primarily through the National Institutes of Health (NIH), is a collaboration between researchers at the University of Michigan, the University of Toronto, Sinai
Medical Center, Harvard University, and the Instituto Nacional de Salud Publica (INSP) in
Cuernavaca, Mexico. Since 1995, project staff have collected and analyzed samples of blood, urine, hair, toenails, breast milk, placentas, and teeth from nearly 7,000 mostly working-class and poor mothers and children recruited through Seguro Social clinics in Mexico City. The bulk of the data collection occurs in a hospital clinic at the Instituto Nacional de Perinatologia (Perinato) during twice-yearly participant visits. The project is managed at the INSP, and at present the bulk of the data analysis occurs at the University of Michigan. Recently the project has expanded to collect data on additional toxins (e.g., mercury and fluoride) and new health concerns (e.g., early sexual maturation and obesity) (Afeiche, et al. 2011).
In the past year I have conducted ethnographic observations of ELEMENT researchers at the University of Michigan in their project meetings, laboratories, and offices, and in Mexico
City, observing data collection with ELEMENT participants. Through interviews and observations with project principal investigators, co-PIs, lab technicians, and project staff, I have mapped the history, organization, and future directions of the project. My preliminary research in both Mexico City and at the University of Michigan has begun to show how the research project operates to generate knowledge and how the ELEMENT study and the life circumstances of the participants exert reciprocal effects on one another.
As an environmental-health study based in global public health, since its inception
ELEMENT has had a relational, organism-environment focus and a more expansive view of exposure than many bioscientific undertakings (Anderson 2004). However, ELEMENT, like most bioscientific research, is underpinned by an assumption of the universality of human biological processes. When ELEMENT aggregates biological sample data from specific bodies in Mexico
City, these data are treated as universally applicable facts .
For instance, one of ELEMENT’s key early findings—that lead exposure has the greatest effect on the developing fetus in the first trimester of pregnancy—has influenced environmental and public health policy around the world, hence “making data global” (Edwards 2010). These generalized findings are directed toward bodies in other places that are assumed to work in the same way (Hu, et al. 2006; Tellez-Rojo, et al. 2004).
My project intends to complicate this view, drawing on work in both the biological and the social sciences that investigates the complex looping effects that organisms and environments have on each other (Hacking 1995; Lock and Nguyen 2010). This scholarship challenges the assumption that biological processes are universal, that is, unaffected by circumstance or history.
Lead levels measured in the blood of people in Mexico City, for instance, might indicate different physiological states and different chemical, environmental, and economic histories and effects than numerically identical levels measured in rural China. Research directions at the ELEMENT project support this approach. Some ELEMENT researchers are explicitly engaged in research
7

Mexican Exposures
Elizabeth F. S Roberts that they see as “post-genetic”: that is, they assume that environmental conditions are powerful forces in shaping organisms. And while ELEMENT’s data have provided valuable findings about chemical exposure, researchers there have also grappled with the fact that chemical exposures do not have predictable, homogeneous effects even among family members living in the same household. To explore this variability, ELEMENT scientists have begun to employ epigenetic analysis in their examination of the health effects of chemical interactions in specific environments in Mexico City. In addition, ELEMENT researchers are interested in my collaboration in order to expand their data and analysis beyond the biochemical and individual level. However, even while some scientists pronounce the end of twentieth-century genetic determinism and proclaim that “Lamarck is back,” the project’s overall approach still emphasizes individuals’ ability to make healthy choices to minimize risk, paying little attention to broader structural realties that constrain human existence and affect bodily realities. I hope that a symmetrical bio-ethnographic examination of the data will challenge this individualistic understanding of ELEMENT participant’s life circumstances.
My pilot research has also identified ways that the ELEMENT project itself affects the research subjects and the phenomena being investigated. My preliminary interviews with
ELEMENT participants in Mexico City, in both clinics and homes, included discussions about environment, disease transmission, and the experience of being long-term participants in the study. Participants described to me how ELEMENT staff diagnosed their children with conditions, like ADHD, that would never have been diagnosed or treated through the Seguro
Social system. The mothers I observed did not appear to experience these diagnoses as alarming.
They possessed a sense of their bodies as malleable by circumstance, exemplifying what historians and anthropologists of Latin America have characterized as a more Lamarckian experience of the body than is common in North America. Indeed, these mothers rarely described ill-health as genetic or as passing from parent to child. Rather, ill-health and disease are transmitted through environmental mechanisms such as neighborhood violence, pollution, and food, and can be alleviated through educational programs or the religious values transmitted through parents.
These mothers also saw their participation in ELEMENT as providing them with resources for overcoming conditions like ADHD and obesity. For participants, who are mostly poor and working-class, being involved in ELEMENT provides forms of attention they do not receive through Seguro Social, such as specialized diagnostic tests and being chauffeured to and from ELEMENT appointments in a private car. Thus, this study population has a specific local history and biology created in part by their participation in the study. New models of organismenvironment interaction suggest that these factors might in turn shape the data collected by
ELEMENT.
Two examples from my preliminary research in Mexico City demonstrate how a bioethnographic approach, including dynamics like kinship, history, gender, material culture, and class, could combine with data collected by ELEMENT to create a fuller account of exposure.
The first involves the diabetes diagnosis of an ELEMENT participant. The second involves
ELEMENT participants’ continued use of leaded dishes.
My interview with Cecelia, a lower-working-class woman, and her two sons (ages 8 and
10) in July 2013 illustrates how we might begin to trace the looping relationships between environments and bodies and articulate them as multiple material, social, and biological exposures. Cecelia had been recently diagnosed with diabetes. Her younger son was microcephalic. Cecelia hadn’t been able to find regular work outside the home, so, besides caring for her sons and the house, she stuffed envelopes at home at piecework rates. Her husband worked as an electrician. Cecelia’s parents lived downstairs, and her mother sold dietary supplements through the company Omnilife. Cecelia showed me these supplements, explaining that she gave them to her sons to help with their pre-diabetes and allergies and her younger son’s mental development, temper tantrums, and hyperactivity. She also told me that ELEMENT had
8

Mexican Exposures
Elizabeth F. S Roberts taught her to give better food to her children and to drink more water than soda.
Soon after I arrived, Cecelia sent the elder son to the tiny convenience store nearby to buy milk for the Nescafé and a packaged snack cake for us to share. This was the only store in the immediate neighborhood, but unlike similar stores in the United States, it sold a fair amount of fresh fruit and vegetables.
Cecelia and her family had moved to this neighborhood a few years earlier. They liked it, but sometimes neighbors teased her younger son, and some were jealous of her husband’s regular employment. This jealousy caused the whole family to suffer from stomach ailments, which she tried to alleviate through the use of herbs and massage. Nevertheless, she didn’t miss her natal village, where angry brujas (witches) would send fireballs to their victims. Cecelia was candid about her own anger problems that caused her to occasionally hit her sons. In the next breath she told me about her family’s patron saint, Anthony de Padua, who guards her sons’ well-being.
Near the end of our conversation, Cecelia described how her entire family’s health was affected by the noise and electromagnetic activity emanating from the enormous power transformer grids that dominated the neighborhood, which in fact is called Alta Tension (High Tension), after the transformers themselves.
What might count as exposures in Cecelia’s life? Through one short interview I learned about the family’s experiences with anger, jealousy, employment, direct-sales firms, dietary supplements, violence, ELEMENT’s educational messages, household saints, insulin, snack cakes, neighborhood conflicts, and electromagnetic radiation. Combining these ethnographic observations with the biological data collected on Cecelia and her family could tell us whether and how these different phenomena might be meaningfully counted as exposures. By examining these exposures in one frame, by positing environments and bodies as participating in looping relationships with varying effects, ELEMENT scientists and I hope to develop a more expansive account of study participants’ lives.
In my preliminary research I also found that even though participant families have learned, through ELEMENT, about the detrimental effects of lead on fetuses and young children, some families continue to use lead-glazed ceramic dishes that were made by their grandmothers before they migrated from the countryside. Family members told me that these dishes were essential, at least on religious holidays, for honoring their grandmothers and as reminders of a different kind of life from their own in urban and sometimes violent neighborhoods. Furthermore, they assert that food cooked in leaded dishes tastes different, literally sweeter. The use of these dishes is part of a larger rejection of the North American imperialism perceived in recent public health campaigns to discourage the use of leaded dishes. An example of this wider response is the rise of restaurants that specialize in cooking and serving food on leaded ceramic dishes. Their advertisements claim that leaded dishes make the food taste better and connect customers to tradition. My preliminary research thus suggests that a bio-ethnographic approach to exposure can generate novel findings about interactions between environment and health.
The ways in which exposures like these condition each other and affect their own uptake are not yet understood, but a symmetrical looping model itself could provide a means for identifying crucial relationships. At ELEMENT, a postdoctoral researcher in epigenetics, Jackie
Goodrich, is interested in learning what a more complex exposure model, which takes into account factors such as household economy, migration history, and gender, might contribute to understanding how diabetes is triggered in ELEMENT participants like Cecelia, or how lead might affect various family members differently. For my part, I plan to develop an approach that includes epigenetic mechanisms along with kinship, life history, economic circumstances and practices, and emotions. The binding proteins that affect blood sugar cannot tell the whole story; neither can parental anger, snack cakes, lead plates, US imperialism, or recent economic policies that reduce the possibility for formal sector employment; but taken together, these exposures can tell us much more about what shapes these families’ bodily and economic fates.
9

Mexican Exposures
Elizabeth F. S Roberts
VI. RESEARCH DESIGN AND SCHEDULE
The proposed project uses a two-phase, multi-sited study centered on a long-term environmental health project investigating the effects of toxins on human health in Mexico City .
Results from my autonomous ethnographic research with families connected to the ELEMENT project will be put in relation to biochemical data collected by ELEMENT researchers. The overall analysis will develop a bio-ethnographic approach to exposures, which, in addition to chemical toxins, includes economic, historical, and experiential domains as well as the scientific research itself. In this study, the ELEMENT project has multiple roles: it is a starting point for ethnographic study (as a means of identifying families and toxins/issues); a collaborator in combining and analyzing data; and an object of study in itself.
PHASE 1 – 2014-15
The primary aim of this phase of the study is to collect data on the environments and circumstances that shape complex states such as obesity and diabetes through ethnographic household research in Mexico City among six working-class and poor ELEMENT participant families. I will focus on three sites: (a) Mexico City as a whole; (b) Barrio Norte, a fully urbanized working-class and poor neighborhood in the central-city Alvaro Obregon delegación; and (c) the working-class barrio of Pedregal de San Nicolás in Tlalpan, Mexico City’s largest and most forested delegación.
To understand the broad history of bodies and environments and exposures studied by
ELEMENT, I will track changes in Mexico City since the beginning of the ELEMENT project in
1995, including reductions in lead levels and other forms of air pollution, increases in obesity rates, and transformations in the health care system, particularly the institution in 2012 of a more inclusive system that covers the unemployed. I will investigate these changes through the publichealth data gathered and disseminated by the INSP in Cuernavaca and the vast Spanish- and
English-language public-health literature about illness and environment in Mexico City, including
ELEMENT publications (Borja-Aburto, et al. 1998; Gómez-Peralesa, et al. 2004; Hernández, et al.
1999).
To gather data on the chemical and relational exposures experienced by specific
ELEMENT participants, I will carry out intensive ethnographic fieldwork in Barrio Norte and
Pedregal de San Nicolás. Both are working-class, poor neighborhoods located near the Seguro
Social clinics where ELEMENT conducted its recruitment efforts. I have selected these two neighborhoods because of differences in their built environment and density, which are thought to affect air quality and lead levels in particulate dust. Barrio Norte is a densely built, hilly neighborhood with full integration of residences and small businesses. Residents characterize it as infested with both crime and garbage but also as tight-knit, with a lively street life. Residents tend to have higher blood lead levels in Barrio Norte than in Pedregal de San Nicolás, located on the edges of a mountainous forest in delegación Tlalpan, which provides 70 percent of Mexico City’s water. A majority of residences have some kind of front yard that many use for informal vending throughout the week. The atmosphere is almost village-like. The participants I interviewed in San
Nicolás felt mostly safe but anxious about the likelihood of increased illegal building and urban expansion into the forests, as well as the influx of drug violence and crime that might come with this expansion.
I will live in these two neighborhoods for six months each. I will recruit the six families
(three in each neighborhood) for my study in consultation with the ELEMENT staff and with the families themselves. At this point, the family characteristics that appear significant to the bodily states I am investigating are family structure (nucleated or extended), age of children, employment status (especially of women), and overall health status. Most ELEMENT participants live in extended-family households with at least three generations of family members, as well as adult siblings and their partners. While this allows them to share resources
(including food, electricity, and money for rent), most family members I interviewed expressed
10

Mexican Exposures
Elizabeth F. S Roberts the hope that one day they would live in more nucleated households. Employment status of women may affect family health: in about half of the participant households, women work primarily outside the home. ELEMENT researchers speculate that in these households, family members eat more packaged and prepared foods, which may increase the rates of obesity and diabetes. However, since the ELEMENT survey design does not collect information about who has the responsibility for preparing food, especially in extended households, this is an issue better addressed ethnographically.
I will conduct life-history interviews and daily participation in and observation of care relations between family members and their neighborhood, the substances they come in contact with and ingest, and the built environments where they live, shop, and travel for work. I will also participate in neighborhood events, such as neighborhood-watch meetings, holidays, and the collective care of neighborhood saints. I will pay particular attention to what family members do when they or other family members are ill, including whom they consult for advice (doctors, pharmacists, neighbors, healers), and where (pharmacy, Seguro Social clinic, private clinic, etc.).
Given my focus on the relationship of life circumstances and ill-health, I will also pay particular attention to economic activity (work, social welfare, vocational training), and its connections to gender and age, and relations within families.
These interviews will be transcribed. Word-for-word interview transcriptions are indispensable for noting and analyzing how concepts, terms, and concerns emerge and recur. I will also code my field notes for recurring patterns. To protect the identities of the participants, each family member will be given a code name, and the interview recordings, transcribed interviews, and raw and coded field notes will all be kept in a secure physical location and electronically stored in Dropbox with password protection.
To aid my research, I will hire a research assistant trained in qualitative research methods and familiar with Mexico City. The assistant will be essential in conducting the complex and time-consuming observations of the daily life of the six families, as well as in conducting archival research and in coding field notes and formal interviews so that they will be nearly complete when I return from Mexico City in July 2015. The remaining notes and interviews will be coded by June 2016. To gain a more longitudinal sense of the six families’ lives, I will rehire this research assistant in 2016–17 to conduct follow-up interviews with key members of each of the families. This data will be invaluable in refining our sense of what counts as a meaningful exposure.
A second aim of this project phase is to study the production and effects of scientific and ethnographic knowledge associated with ELEMENT’s project and to integrate this knowledge into a larger analysis of environments, exposure, and bodily states in Mexico City. I will engage with the ELEMENT study by a) observing the ELEMENT testing facilities at the Perinato; b) observing and accompanying the six chosen ELEMENT participant families before, during, and after their twice-yearly ELEMENT appointments; and c) giving the six ELEMENT participant families grocery-store gift cards and observing family members shopping with the cards.
The Perinato is where ELEMENT staff conduct most of their medical, neurobehavioral, and developmental assessments and collect biological samples from study mothers and children.
In addition, ELEMENT staff assess the participants’ home environments through home visits.
During 2014–15, I will spend one day a week with the Perinato staff. In the mornings I will observe staff and study participants interacting in appointments. In the afternoons I will accompany staff on home visits. I will also observe the introduction of a new component of the
ELEMENT project, which will compare archived biological samples with environmental samples
(like water, table salt, and toothpaste) from participant households in order to assess the impact of childhood exposure to fluoride on cognitive function.
To understand how participation in ELEMENT itself shapes individuals’ life circumstances and how they are constituted as test subjects, I will accompany the mothers and children from each of the six families on their twice-yearly ELEMENT visits to the Perinato.
11

Mexican Exposures
Elizabeth F. S Roberts
Observing these visits, the required preparations (e.g., fasting before blood tests), and the followup activities (e.g., saliva sampling and keeping activity journals) will provide me with a participant’s-eye perspective different from my role as an observer alongside ELEMENT staff.
Additionally, I will ride with the participants to and from the Perinato in the private car
ELEMENT sends for their transport—a mode of travel that participants describe as usually reserved for the rich. As a form of personal attention that both solidifies participants’ commitment to ELEMENT and materially shapes their life trajectories, this experience exemplifies the type of looping effect I plan to study.
In order to learn more about the relationship of food consumption and family relations, I plan to give each family five $100 grocery-store gift cards over five months. (ELEMENT already offers grocery gift cards in much smaller amounts, so participants are accustomed to this type of exchange.) I am hoping most families will allow me to come to the grocery store when they use the cards so I can talk with family members about what they are buying and why and observe the gendered relations of care involved in shopping and cooking for families, while also documenting the effect of the card and my own observations on their purchases and food preparations.
PHASE 2 – 2015-17
Following my fieldwork, I will conduct two years of research focused on the ELEMENT research project in Michigan, with targeted trips to Mexico City and ELEMENT’s offices at the
INSP in Cuernavaca, Mexico, primarily in the summers .
This phase of the project will continue the study of the production and circulation of scientific knowledge. It will examine how ELEMENT researchers codify and measure exposure and how the project transforms biological data from specific test subjects embedded in particular environments into generalized knowledge about “universal” bodies. I will use interviews and participant observation in ELEMENT research laboratories and group meetings to further map the formation and circulation of data, concepts, and findings about the relation of toxins to bodies and health in Mexico City.
This phase will be critical to the third aim of my project: to develop a bio-ethnographic approach that will expand the parameters of “exposure” by incorporating both biological data (from the ELEMENT team) and my ethnographic data.
At this preliminary stage, ELEMENT researchers and I plan to link my data about household economies, gendered and economic practices of food preparation and consumption, kin relations, built environments, and urban infrastructures with
ELEMENT data relating lead exposure to physiological mechanisms that might trigger obesity and diabetes.
Part of our effort will involve discussions of the definitions of exposure and susceptibility to exposure. The concept of exposure could encompass not only specific chemical substances
(like lead) but also conditions such as poverty, family dynamics, violence, gender relations, economic circumstances, and the ELEMENT study itself. By understanding exposures as openended conditions that continually interact with participants, we can ask how they affect individuals’ susceptibility to forms of ill-health like diabetes. Additionally, this approach will allow me to document ethnographically how exposure is defined by ELEMENT scientists, study participants, and myself.
To understand the variability of their data, ELEMENT researchers are interested in disaggregating their large data sets on lead exposure to examine the data collected on the six families I study. This process of disaggregation is in a sense the opposite of what these researchers usually do, which is to aggregate individual data to make generalized claims. While much of the specific agenda for this phase of the project will be developed on the basis of the ethnographic data I collect in 2014–15, my interviews from July 2013 have suggested one possible approach. As described above, we will associate my ethnographic data about participants’ exposure to lead in food dishes and about gendered and intergenerational care dynamics in specific households with the numerical data on the blood lead levels of family
12

Mexican Exposures
Elizabeth F. S Roberts members over the last 18 years, then compare these results with those from another of the six participant households who no longer use leaded dishes. The experimental objective of this collaborative analysis will be to see what kinds of questions we can ask, and begin to answer, if lead levels, migration histories, and gender dynamics are all understood as equally real, constructed, and contingent.
To facilitate the analysis of this bio-ethnographic data, I will hire a PhD student in medical anthropology at the University of Michigan. Our data analysis will be undertaken in consultation with ELEMENT researchers at the INSP in Mexico, including visits there in 2016 and 2017 to work with Dr. Mara Tellez-Rojo, ELEMENT director at INSP, and her staff.
VII. QUALIFICATIONS, PREPARATION, AND PERMISSIONS
My investigative approach in Mexican Exposures is shaped by my previous ethnographic studies of science and medicine, particularly in Latin America. I am trained as a medical anthropologist of Latin America (PhD, UC Berkeley/UCSF Joint Program in Medical
Anthropology, 2006). I am a seasoned fieldworker, having conducted research in the United
States, Ecuador, and Mexico City, as well as at international conferences and events. In Latin
America, I conducted all ethnographic observations and interviews in Spanish. My research has explored the historical and contingent dimensions of bodies across several projects. In articles and an ethnographic monograph, God’s Laboratory: Assisted Reproduction in The Andes (Roberts
2012), I analyze the relation of nature and culture, science and religion as they are manifested through in vitro fertilization (IVF) clinics in contemporary Ecuador, a primarily Catholic nation.
Drawing on the semiotic-materialist approaches of science studies, medical anthropology, and feminist scholarship, my research thus far has demonstrated the malleability of the biological realm in the practice of IVF in Ecuador; analyzed the material effects of place on bodies and bodily processes (Roberts 2013); and developed a model for understanding how broad regional political and economic shifts have made reproduction such a complex focal point for politics in
Latin America (Morgan and Roberts 2012). I have also used my teaching to deepen my knowledge of the relevant anthropological and STS literatures for this project, including coverage of intergenerational transmission of disease, kinship, and public health. My findings in Ecuador on the malleability of race and nature shaped my current interest in examining how both scientists and laypeople experience “environment,” broadly conceived, as shaping human bodies and human disease and health.
In addition, I have experience in collaborating with colleagues from medicine and science. My Ecuador IVF project involved developing literacy in the biology of human reproduction and techniques of reproductive technology. Developing and team-teaching a course with a biological anthropologist on the nature/culture dichotomy in anthropology and everyday life has allowed me to explore the intersection and divides between biological and cultural approaches in a pedagogical context. In 2011–12, a K-12, NIH–BIRCWH grant (Building
Interdisciplinary Research Careers in Women’s Health) underwrote my study of the burgeoning field of epigenetics and environmental health through formal coursework at the University of
Michigan School of Public Health, in interviews and observations of environmental health and epigenetic scientists, and by participating in relevant conferences around the United States.
Having worked with the ELEMENT project for the last year, I have developed the appropriate support and permissions for this project. Project investigators have welcomed my engagement in their meetings and observations of their project at the University of Michigan and in Mexico City and Cuernavaca. In 2013, ELEMENT staff in Mexico City recruited participants in their study for my pilot observations and interviews. ELEMENT PIs are enthusiastic about the possibilities of combining my ethnographic data with their biological data in order to develop new questions and methods for assessing exposures. I include here a letter of support from Dr.
Karen Petersen, Professor of Environmental Health Sciences and Nutrition in the University of
Michigan School of Public Health and co-PI of the ELEMENT project. I am already in
13

Mexican Exposures
Elizabeth F. S Roberts conversation with Dr. Petersen and Dr. Mara Tellez-Rojo (ELEMENT director at the INSP in
Cuernavaca, Mexico) and relevant staff about seeking IRB approval for my project from the
University of Michigan, the NIH, and the INSP, a process that should be completed by June 2014.
Dr. Tellez-Rojo will also assist me with securing my visa to live in Mexico for the year.
VIII. INTELLECTUAL MERIT
Mexican Exposures will model the application of recent STS and anthropological critiques of the nature/culture divide through a specific case study that considers a range of material, social, and experiential influences beyond chemicals, in ways that can capture the potential looping effects of interrelated entities. My development of this method and subsequent publications will contribute to the growing anthropological and STS literature that rejects nature and culture as a priori explanatory devices and instead foregrounds the contingent histories of human lives, bodies, and biologies (Fuentes 2010; Singh 2012). In particular, Mexican Exposures will investigate the specificity of the linkages between chemical exposures and the relational worlds of families in Mexico City without assigning either chemicals or care relations to nature or culture.
Currently, ELEMENT researchers are investigating how individual chemical exposures might trigger epigenetic mechanisms that govern physiological growth, attention and sexual maturation and may contribute to developmental delays, diabetes, and obesity. A typical ethnographic approach to these issues would focus instead solely on social and historical issues, leaving the understanding of the physical effects of lead exposure to biological scientists. That division of labor, however, separates chemical substances, on the one hand, and ideas and practices, on the other, into the discrete domains of nature and culture. Within this scenario, lead is seen as a constant, a chemical element with predictable biochemical effects. By contrast,
Mexican Exposures will integrate the quantitative data on blood lead levels with an examination of how “social” practice and history might affect the presence of lead in particular bodies, assuming both lead levels and practice are contingent. By linking their biological data from study participants with my ethnographic data about how and why leaded dishes are used, and by whom, we will put ethnographic and biological exposure data in a single frame. Thus a bio-ethnographic approach could trace how specific histories of rural-urban migration, the taste of traditional foods, and specific gendered intergenerational care relations complicate the conditions of individuals’ exposure to lead, and blood lead levels themselves , telling a more complete story than blood samples alone could provide. By linking family members’ blood lead levels with their family relational histories, this approach takes biological data seriously but precludes such data from being viewed as representing the “real” exposure story or the universal story of lead everywhere.
Through this project I will produce an ethnographic monograph and several articles for publication in anthropological and STS journals like American Anthropologist , American
Ethnologist , BioSocieties, and Social Studies of Science . This work will have two dimensions: 1) tracing the contingent local biologies of study participants. and 2) analyzing how scientific knowledge (including my own) about exposure and health is produced in relation to specific bodies and environments. These publications will make several contributions to the larger field of anthropology by describing an empirical project that decenters the nature/culture dichotomy as a means for understanding the relationship of life circumstances to disease; by placing emergent models of health and environment produced by the ELEMENT project in conversation with nature/culture debates in twentieth-century anthropology; and, most crucially, advancing a bioethnographic approach to the complex interplay between diverse human physiologies and the historical, political, economic, chemical, and biological exposures that contribute to disease. By refusing to reify culture and nature as separate domains, Mexican Exposures will also contribute to the contentious anthropology and social science and public policy debates on the “culture of poverty.”
14

Mexican Exposures
Elizabeth F. S Roberts
IX. BROADER IMPACTS
The impact of Mexican Exposures will extend beyond anthropology and STS through joint publications with ELEMENT researchers in interdisciplinary, environmental health, and global health journals like Environmental Health Perspectives and Science of the Total
Environment , as well as journals like Global Public Health . These publications will focus on our methodology and the findings produced through our collaboration. The purpose of these publications will be twofold: to create a framework for exposure that moves beyond seeing single environmental or social determinants as primary mechanisms for disease, and to use a bioethnographic approach in conjunction with the findings of epigenetics and evolutionary developmental biology to question how the complexity of ill-health, as manifested in specific chemical relational exposure environments, is generalized for application to health policy worldwide. Through such publications, we hope to provide a way of making useful interventions in the global health policy literature that currently prioritizes big data, drawn from aggregate and generalized data sets, by challenging the assumption that biological processes are universal—that is, unaffected by specific life circumstances and history (Erikson 2012).
Although the biological and social sciences are reconfiguring assumptions about what constitutes their objects of study, this reconfiguration is still in an early stage. For the most part researchers continue to assume a difference between nature and culture and to focus on one or the other. Mexican Exposures , by contrast, does not regard nature and culture as distinct domains.
Instead, I am proposing an experiment that brings two methods together with a shared understanding of both social and biological processes as constructed and contingent. We will incorporate biological and chemical processes as well as human behavior and history into our collaborative understanding of the conditions that produce ill-health. I hope that the bioethnographic approach I seek to develop will provide tools for research and analysis that do not rely on a nature/culture divide. Ultimately, Mexican Exposures will succeed if it produces innovative questions and more complex accounts of exposure and the relationship between disease and life circumstances than would be possible through either ethnography or environmental health science alone.
15